Changelog

0.1

  • Reorganization of plotting interface: input parameters are now more consistent across different plotting functions, default parameters changed

  • Set default colors for all the plots

  • Add distribution colors and colorblind safe option for high dimensional plots

  • Bug fix in default computation of ranges in contour plots

  • Dynamic computation of ranges based on samples

  • Support for time series analysis:

    • Time series plotting functions

    • Uncertainty-aware seasonal trend decomposition based on loess (UASTL)

    • Uncertainty-aware Fourier analysis

  • Use pytest for regression tests

  • Add Shepard plots for quality evaluation

0.0.2

  • Reorganization of imports

  • Documentation of distribution class

  • Hideable mean in violin plot

  • Consistency of various plotting functions

    • plotting functions return created figure and axis objects

    • plotting functions have optional parameter showplot resulting in display of the figure

    • same parameter naming

  • controllable width for box and violin plot

  • dedicated plotting functions for 1D distributions plots e.g. generate_boxplot(...)

0.0.1

Features in first version:

  • Uncertainty-aware Principle Component Analysis

  • Uncertainty-aware Multi-dimensional Scaling

  • Basic visualizations:

    • N-dimensional distributions: scatterplot matrices, iso-lines of probability distribution

    • 1D distributions: box plots, violin plots, dot plots